Scaling rights: the ‘Turkey deal’ and the divided geographies of European responsibility

ABSTRACT In this article, Bialasiewicz and Maessen examine how responses of the European Union (EU) to the refugee crisis have been differentially spatialized within and outside the EU’s boundaries, noting how the crisis has operated a geographical sorting, not just of the right to legal and humanitarian protection, but also of the right to be included within the spaces of EUrope’s presumed responsibility. They highlight, in particular, the divided affective geographies delimiting concerns with ensuring the bodily safety of Europeans within the EU’s member states from the need to ensure safe passage for refugees at and beyond the EU’s borders. Such divided geographies made themselves violently visible in the spring of 2016. As EU politicians from the right, centre and left were calling with seemingly one voice for the need to assure ‘the protection of European women’ from what was being presented as an unprecedented surge of sexual attacks perpetrated by newly arrived ‘refugees’, at that very same moment, fundamental legal notions of safety and protection were being rescripted as part of the EU’s ongoing negotiations with Turkey to take on the management of refugee flows at the EU’s external borders. In the current piece, Bialasiewicz and Maessen focus on the ‘Turkey deal’ specifically, but they also locate these events in a broader re-scaling of EU responsibilities over the past decade, noting the effects of such re-scaling on the access to basic rights: within, outside and at the borders of Europe.

As a contribution to the discussion of the imagined temporalities of the refugee crisis in this special issue, we wish to draw attention also to its imagined spatialities, noting how the two function in concert to produce particular understandings of what EUrope is, whom it should care for and where. 1 We focus on the so-called 'Turkey deal', announced in March 2016 in response to mass arrivals of migrants at the European Union's (EU) southeastern borders. The 'deal' is illustrative, we argue, because it highlights in stark (and violent) fashion the spatial and political ambiguities that characterize contemporary EU border and migration management. By looking at how responses to the refugee crisis have been differentially spatialized inside and outside the EU's boundaries, we reflect on how the crisis has operated a geographical sorting not just of the right to legal and humanitarian protection, but also of the right to be included in the memory of 'what Europe stands for', and within the space of EUrope's responsibility.
As colleagues from the New Keywords Collective, writing on 'Europe/ Crisis' in early 2016 reminded us, there is a pressing need to unsettle radically both the conceptual categories in which we seek to understand the European discourse on 'migration' and 'refugees', and its presumed geographies, in particular contesting its imagining as a phenomenon 'external' to (what we take to be) Europe. 2 We wish to take up their call by focusing on some of the forms of externalization of the refugee crisis that have consented at once a distinct migrant politics within EU member states, and a distinct geopolitics in relation to the EU's neighbours. In our analysis, we look specifically to how spatial separation and distanciation have created divided affective geographies delimiting concerns with ensuring the bodily safety of Europeans within the EU's member states from the need to ensure safe passage for refugees at and beyond the EU's borders: divided affective geographies that separate out the subjects of EUrope's care and (also historical) responsibility.
When the President of the European Commission Jean-Claude Juncker delivered the State of the Union address on 9 September 2015, right at the apex of the refugee crisis, it was precisely an appeal to a common European historical responsibility and 'historical fairness' that framed his call to the assembled members of the European Parliament: … for Europe it is also a matter of historical fairness. We Europeans should remember well that Europe is a continent where nearly everyone has at one time been a refugee. Our common history is marked by millions of Europeans fleeing from religious or political persecution, from war, dictatorship, or oppression. … We Europeans should know and should never forget why giving refuge and complying with the fundamental right to asylum is so important. … I am counting on you, in this House, and on all Member States to show European courage going forward, in line with our common values and our history. 3 In his speech, Juncker was careful not to specify to whom and where EUropean refuge was due: indeed, he noted that Europeans should be proud, not fearful, 2 New Keywords Collective, 'Europe/crisis: new keywords of "the crisis" in and of of the fact that 'today it is Europe that is sought as a place of refuge and exile. It is Europe today that represents a beacon of hope, a haven of stability in the eyes of women and men in the Middle East and in Africa.' 4 Yet this unifying appeal to EUrope's supposed legacy of refuge remained only that: as one commentator surmised, 'while Juncker's speech in the European Parliament was strong on history, it fell flat when it came to policy proposals in this crisis'. 5 Indeed, in the weeks and months that followed, the EU became increasingly geographically divided in its response to the crisis, with also a geographical sorting of the right to Europe's legacy of 'historical fairness'.
These divided geographies made themselves violently visible in the spring of 2016. As EU politicians from the right, centre and left were calling with seemingly one voice for the need to assure 'the protection of European women' from what was being presented as an unprecedented surge of sexual attacks perpetrated by newly arrived 'refugees', 6 at that very same moment, fundamental legal notions of safety and protection were being rescripted as part of the EU's ongoing negotiations with Turkey to take on the management of refugee flows at the EU's external borders. In the current piece, we focus on this particular political moment, and on the 'Turkey deal' specifically, but it is important to locate these events in a broader re-scaling of EU rights and responsibilities over the past decade.
Within the EU, scholars have long noted the emergence of differential 'affective geographies', sorting populations and individuals into those to be feared and those to be cared for. 7 In their introduction to the recent special issue of Patterns of Prejudice on 'Anxious Politics in the European City', Wayne Modest and Anouk de Koning stress how the emergence of new affective geographies of anxiety in Europe cannot be separated from wider social, political and economic shifts occurring over the past several decades in European welfare states, shifts related to broader forms of political-economic restructuring (and crisis), as well as to demographic transformations (both real and perceived) in European populations. As 'European countries struggled to come to terms with the social changes in the make-up of welfare states and national populations', they note, 'this led to the emergence of new questions of how to care for one's citizens or even who should be understood as a citizen to be cared for ' The question of 'who should be understood as a citizen to be cared for' is fundamental, as it speaks to both actual juridical-political definitions of (and boundaries to) citizenship, as well as to the wider affective geographies inscribing who belongs to the community 'to be cared for'; wider affective geographies inscribing who should be the subject of care and concern, and of those much-abused terms 'European solidarity' and 'responsibility'. Modest and de Koning emphasize, in fact, that the creation of hierarchies of belonging to the national (and European) community is being directly tied today also to a spatial differentiation of rights, 9 and this is the crucial point that we wish to pick up on here.
Again, such forms of differentiation are in no way new to the refugee crisis of 2015 and historical parallels could be traced with various other instances in Europe's not-that-distant past, when the sorting of refugees' bodies and rights was so powerfully described by Hannah Arendt. 10 Nevertheless, the past two decades have witnessed a growing decoupling of political as well as socioeconomic rights from territory within EU member states and, all the while, their increasingly territorialized policing. 11 Whether it is attempts to specify differential rights of intra-EU migrants to welfare benefits (suggested, or already adopted, by a number of states) or proposals for the outright stripping of basic political rights, including citizenship, from those considered a security threat (most prominently discussed in France but also in a number of other EU states following the Paris attacks), forms of what Ranabir Sammadar terms 'differentiated citizenship' have become evermore pervasive. 12 While such differentiation and deterritorialization of the access to rights is gaining ground (and acceptability) within EU member states, it has been even starker at the EU's borders and in the EU's more extended forms of 'border-work'. As we have argued elsewhere, 13 for over a decade now, the 'border-work' of EU member states has stretched far beyond the physical borders of the EU. Border officials now routinely speak of 'remote control borders' (or 'bordering at a distance'), of 'layered inspection strategies' or of the 'thickening of borders' through 'buffer-zones'. The EU's neighbours to the east and south have played a crucial role in these policies, whether through EU-financed hardening of controls at their own borders with states further afield (such as through the EUBAM programme in Ukraine and Moldova, launched in 2005), or by participating in joint policing operations designed to prevent migrants from reaching the EU's external borders (as has been the case in the Mediterranean over the past decade). Many such agreements have been bilateral in nature, directly engaging states such as Spain and Italy with North African counterparts (Morocco in the case of the former and Libya in the case of the latter). 14 The differential de-bordering of the EU's borders has not only relied on such direct collaboration with more or less unsavoury regimes at its external borders, charged with halting irregular migration flows. The control of the EU's borders has also progressively been transformed into a highly complex and geographically dispersed system of 'border management', relying on a much wider array of institutional and private actors, far beyond the territory of the EU. Most often, the first 'border guard' that a traveller to the Schengen space will encounter is indeed the travel agent or airline company representative, now tasked with assessing their immigration and visa status (Canadian political theorist William Walters has coined the term 'viapolitics' to refer to the crucial role that modes of transport and transport routes play in power struggles over mobility, and especially over its control: very much a crucial form of power in the attempts to halt the flows of refugees across the 'Balkan route' in the autumn of 2015). 15 The role of external actors, state as well as private, in stemming but also (and equally importantly) in sorting and administering migrant flows, has become crucial to the EU's border-work, together with the proliferation of off-shore reception and detention centres and 'regional protection' facilities. A key function of such facilities has been to act as 'sorting locations' in the management of migration flows: as various scholars have argued, they are spaces in which the right to certain rights is determined, where access to citizenship is (differentially) deliberated. 16 This is a crucial transformation of the border regime, for the centres and camps that already exist at the EU's borders (and those being proposed) are not simply deterritorialized, exceptional, 'waiting spaces' in which European rights do not (yet) apply. They are rather sites that are crucial to the sorting and organization of the right to European rights, through a principle of differentiated inclusion. Access to the right to asylum in the EU is thus no longer regulated through physical presence on the territory of an EU member state, but determined in geographically dispersed locations.
There is another important set of geographies at play in the EU's borderwork, and this too is in no way novel to the crisis of 2015. Together with the off-shoring of such sorting and detention locations, the past years have also seen an important shift in the political discourse on responsibility for migration. It is therefore not just the physical management of migration that has been externalized, but also the 'blame' and attribution of responsibility. As Maribel Casas-Cortes et al. argue in their analysis of the dynamics of externalization, 17 irregular migration has long been presented by EU institutions as the direct product of the failures of states in the European neighbourhood to secure their territories and borders properly. So too with the crisis of 2015, presented as 'inflicted on Europe' by external agents whose responsibility it therefore was to solve the crisis. 18 The 'Turkey deal' announced in March 2016 could be seen as simply part of a variety of forms of externalization and distanciation of the responsibility for EUrope's migration management, also since it aimed to replicate alreadyexisting agreements with the EU's other neighbours to the south and east. Nevertheless, it merits our close attention, both for its scale and ambition, but especially for its explicit suspension and rescripting of EU member states' obligations under the 1951 Geneva Convention on the Rights of Refugees, in particular the obligation of non-refoulement (push-back) to countries not considered 'safe'. It is also revealing of the divided geographies of EUrope's 'historical responsibility', so frequently invoked in the months that preceded the agreement.
The 'deal' and its divided geographies Faced with an 'unmanageable' flow of people through the 'Balkan route' into the EU, in the early spring of 2016 a proposal was brought forward by German and Dutch politicians to 'restore control over the external border in the Aegean'. 19  all migrants reaching Greece after the date of 20 March 2016 would be returned to Turkey; once numbers of arrivals had fallen, Syrian refugees specifically would be resettled (on a voluntary basis) to EU states. The 'deal', however, was presented not simply as a necessary solution to regiment the 'chaos' of unregulated mass migration flows at the EU's southern borders. Equally important in justifying its terms, as put forward by the deal's Dutch and German proponents, was the need to 'combat the business model of smugglers' by ensuring the safety of migrants attempting irregular crossings into the EU: migrants' forcible return to Turkey would, in the deal's terms, 'offer [them] an alternative to putting their lives at risk'. 20 With the Netherlands holding the EU Presidency in the first semester of 2016, Dutch politicians were key in preparing the ground for the 'deal' (most prominently, Prime Minister Mark Rutte and the then-leader of the Dutch Labour Party (PvdA) and Rutte's coalition partner, Diederik Samsom, after whom the deal is named). Just days before the proposal was presented, PM Rutte suggested that the deal would certainly 'decrease the inflow of refugees to zero'. 21 Claims that the 'deal' would 'protect' migrants were fundamental to the case made by the Dutch, with the proposal's co-sponsor Samsom arguing forcefully that the deal would entirely eliminate the incentive for refugees to make the dangerous crossing from the Turkish Aegean coast to the Greek islands. Samsom argued, moreover, that despite the objections of the office of the United Nations High Commissioner for Refugees (UNHCR) and other organizations such as Amnesty International that condemned the 'deal' as illegal and inhumane, 22 returning refugees would be fully in agreement with international law, as long as procedures were executed 'properly and with care'. 23 In order to make the deal possible, however, Greece would be required to apply the principle of 'safe third country' (a fundamental pillar of the EU Asylum Procedures Directive, Article 38) to Turkey, enabling it to return to the latter's territory refugees who crossed the Mediterranean from Turkey to Greece. 24 Importantly in these debates, a distinction was made between the nature of that 'safety', that is, between a 'safe third country' for refugees to now be returned and a 'safe country of origin'. Under the EU Asylum Directive, a 'safe country of origin' is considered safe for its own citizens (defined as free from persecution or bodily harm), while a 'safe third country' is deemed to be 'safe' for citizens (and, more specifically, refugees) of other nationalities. We will say more about how this distinction was scripted and operationalized in the 'deal' in subsequent sections of the paper, so we simply highlight here how it crucially relied on a differentiation of the right to 'safety' and protection. Another rubric of differentiation was key to the arguments made for the 'deal' as well. Its EU proponents were in no way preoccupied with hiding their thoughts on their prospective partner. Suggestions that Turkey might gain leverage over the EU through the 'refugee deal' that was part of the negotiations were directly dismissed. 25 PM Rutte argued that, dealing with Turkey was necessary, regardless of the domestic situation in that country, because 'if you only want to deal with countries that hold the same standards with regard to press freedom and human rights as ourselves, then we will turn out to be very lonely'. 26 To 'protect Europe' from the chaos of unregulated migration flows-and refugees themselves from unscrupulous smugglers and the dangers of land and sea crossings-EU standards and the rights enshrined within EU and international legislation had to be re-assessed. 27 24 'Turkey as a "safe third country for Greece"', 17 October 2015, available on the European Stability Initiative website at www.esiweb.org/pdf/ESI%20-%20Turkey%20as%20a% 20safe%20third%20country%20-%2017%20October%202015.pdf Appealing to the exceptional, 'crisis' nature of that particular moment, EUropean responsibility (and legal obligations) could thus be suspended, shifting both the task and the moral and legal register elsewhere. The protection of EUrope-but also, ostensibly, of the migrants themselves ('from unscrupulous smugglers')-required a divided security regime. It is on this divide that we now wish to focus, stressing how such an unequal operation of security regimes constitutes a powerful political and geopolitical practice, deciding and delimiting who is to be kept safe, and where.
Over the past decade, scholars have noted how the War on Terror has been sustained by strictly bounded geographical imaginations of 'safety', invoked to justify both military interventions abroad, and the excesses of the security state at home, both in North America and in the EU. 28 A geographically differentiated 'logic of protection' has been fundamental in sustaining such interventions and excesses: as Iris Marion Young argued in her 2003 essay, 'the Bush administration repeatedly appealed to the primacy of its role as protector of innocent citizens and liberator of women and children to justify consolidating and centralizing executive power at home and dominative war abroad'. 29 This logic of protection relied on a series of gendered bodily and geographical distinctions: 'constituting the "good" men who protect their women and children by relation to other "bad" men liable to attack'. 30 As feminist geographers and scholars in critical security studies have noted, 31 that which Lila Abu-Lughod termed 'a new colonial feminism' 32 has been key in the tracing of distinct 'myths of protection', both at home and abroad, constituting subjects in need of protection, threatening others and 'securitizing actors'. 33 Indeed, the 2015 refugee crisis, at first presented as a crisis of/at EUrope's borders, very quickly shifted in both political discourse as well as material response from simply a question of 'external' border management to a question of maintaining 'internal' order. The need to maintain 'order' did not only regard member states' physical and institutional 'carrying' and 'reception' capacity. Faced with a 'wave' of hundreds of thousands of mostly young male refugees, the popular and political rhetoric of the catalogue of potential challenges represented by large-scale irregular migration quickly expanded its scope of dangers to fundamental risks to the EU's very stability and social order. The events in Cologne and other European cities on New Year's Eve 2015 became an unfortunate turning point in these discussions: the refugee-as-potential-terrorist became rescripted also as the refugee as potential rapist, molester and abductor of European women, with politicians from Austria to Germany to the Netherlands all calling with urgency to preserve the safety of women's bodies from the impending threat posed by the 'refugees' and 'asylum-seekers' (the terms used interchangeably) now roaming the streets of European cities.
From tabloid newspaper (and not only) covers depicting (white) female bodies attacked by dark groping hands, to public initiatives to arm women with self-defence spray by far-right parties such as the PVV in the Netherlands or 'Team Stronach', a right-nationalist fraction in the Austrian parliament, 34 such framings became part of a wider narrative, with the female body under attack becoming the embodiment of an attack on Europe itself, on its 'values', its 'social rules' and its 'civilization'. 35 The 'rescue narratives' invoked in the spring of 2016 appealing to 'saving white women from brown men' in EUrope's cities echoed the very 'colonial feminist' imaginations described by Young and Judith Butler, with precisely the very same sort of geographical and bodily sorting of the right to 33  protection, 36 a right that has been quite differently refracted in the EU's attempts to 'deal' with the migrant crisis beyond its borders.

Making Turkey 'safe'
While the popular press in Austria, Germany and the Netherlands was constructing these narratives of 'rescue' and appeals to protection in the early spring of 2016, a parallel set of discussions was taking place on a different question of 'safety', with the very same three EU states as key protagonists.
Following several years of increasingly strained relations between the EU and Turkey, and with Turkish EU-accession negotiations practically at a standstill, the refugee crisis of 2015 brought a sudden 're-set': another of the crisis' 'productive' effects. A key moment in this regard was President Recep Tayyip Erdoğan's visit to Brussels, followed by visits of the European Commission's (Dutch) Vice-President Frans Timmermans and German Chancellor Angela Merkel to Turkey in October 2015. 37 The timing of these visits was striking given the Turkish domestic context. Peace and disarmament negotiations between the Turkish state and the PKK (Kurdistan Workers' Party) had been controversially broken off following national elections in June 2015.
The election results had been dramatic for the ruling AKP (Justice and Development Party) since they failed to gain a majority in parliament and witnessed, for the first time, the left-wing HDP (Peoples' Democratic Party) acquiring a solid presence. The successful formation of a coalition government between the AKP and any of the other parties was rumoured to have been obstructed by President Erdoğan who allegedly preferred a majority government, leading to a snap election later that year. 38 These and other worrisome political developments had, in the preceding months, led EU officials to stress repeatedly their concern with the domestic situation in the country. Nevertheless, despite such widely expressed concerns, the European Commission in October 2015 announced a delay in its long-awaited annual progress report on Turkey. EU officials and press representatives repeatedly denied that this had anything to do with the visit of President Erdoğan to Brussels, or the negotiations regarding a 'migration action plan'. 39 The minutes from a meeting between President Erdoğan, the Turkish Foreign Minister Feridun Sinirlioğlu, EU President Donald Tusk and EU Commission President Jean-Claude Juncker indicate, however, that the report was delayed at Erdoğan's direct request. 40 Both the controversial delay of the report, as well as the accelerated exchange of visits of top officials, suggested that the EU was willing to close both eyes to the excesses of Erdoğan's authoritarianism in the interest of a greater 'security good': the need to contain the chaos of unregulated migration on the EU's southern flank. Again, the logic of 'crisis' proved 'productive' in allowing for the suspension of EU oversight of Turkey. The findings of the delayed report were indeed of grave concern: outlining, for instance, the severe political pressure to which the judicial system had been subjected, quoting involuntary transfer of judges and prosecutors, and the government's gradual undermining of the judicial system's credibility by accusing judges and prosecutors of being part of a 'parallel structure'. 41 In those same months, the Turkish government had indeed grown increasingly insensitive to EU criticism in the field of human rights and freedom of speech, with an intensification of judicial and physical attacks against its domestic enemies, notably the AKP's former ally, the Gülen movement ('the Fethullahist Terror Structure' or FETÖ in government discourse) and the PKK. President Erdoğan included also academics in his framing of state enemies and declared that the citizenship of 'accomplices' of terrorists should be revoked. After a petition was launched by scholars asking their colleagues to speak out against state violence in the southeast of the country, Erdoğan vowed to prosecute the signatories. 43 Considering all of the above violations -in flagrant disregard for EU-mandated standards of human rights and liberal democracy-it is all the more remarkable that EU officials pushed towards a recognition of Turkey as a safe third country through the course of the autumn of 2015, as part of the rescripting of a broader readmission and visa agreement between Turkey and the EU.
In some aspects, the 2016 deal defining Turkey as a 'safe third country' for asylum-seekers was part of much longer negotiations between Turkey and EU institutions, aimed at pushing back irregular migrants (namely, those who do not qualify for refugee status) out of the Schengen area and towards Turkey. What was new in the 2016 deal, however, was the possibility for refugees to be pushed back from EU territory before their claims and rights to asylum in the EU were fully assessed and processed. 44 To render that possible, Turkey was now redefined as a safe third country for Greece, a redefinition that was judicially highly problematic. As UNHCR has argued, Turkey does not meet the requirements of Article 38, clause e, of the Asylum Procedures Directive: 'the possibility exists to request refugee status and, if found to be a refugee, to receive protection in accordance with the Geneva Convention.' 45 Rather than granting refugee status to Syrian asylum seekers, for instance, Turkey is only willing to grant them 'temporary protection' under the Turkish Law on Foreigners and International Protection. 46 Negotiations with the EU on readmission had been a crucial part of Turkey's accession process even before the negotiation talks were officially opened in 2005. Alexander Bürgin indicates that Turkey had previously been reluctant to discuss a readmission agreement since the Turkish government was wary that Turkey would become a final destination for irregular migrants, rather than a point of transition. 47 The European Commission, however, made readmission talks a precondition to starting membership talks. 48 Turkey in the end agreed to begin readmission talks, which were obstructed in 2006 due to preconditions that Turkey wanted to have in place before the agreement was implemented, as well as the Commission's decision to block eight chapters in the negotiation process and the decision of the Republic of Cyprus to block the opening of new chapters in 2009. 49 Nevertheless, over time the Turkish government became reluctant to implement policies that were of no direct benefit to Turkey. The Commission therefore tried to create an incentive by offering visa liberalization in return for renewed talks on a readmission agreement in 2012. 50 As Burç Aka and Nergiz Özkural show, Turkey made it clear in 2012 that a readmission agreement would only be signed in parallel with the official start of talks on visa liberalization of Turkish citizens travelling to the EU. The conclusion of this process followed in December 2013, with the initiation of the visa liberalization negotiations and the signing of a readmission agreement, which obliged Turkey to readmit third-country nationals and stateless persons, but only three years after the whole agreement would go into effect. 51 The basis for the current 'deal', in which an accelerated visa agreement featured prominently, was the previously mentioned Merkel-Samsom Plan. An important prerequisite of this plan was that Greece would designate Turkey as a 'safe third country', 52 a shift in geographical representation that would consent a geographical redirection of the migrant flows. To lay the ground, a differentiation between the concepts of 'safe country of origin' and 'safe third country' (for refugees) had to be specified. 53  by the EU deal-makers was precisely the fact that Turkey was not considered a 'safe country' for refugees because of the multiple violations of Turkish citizens' own human rights. 54 Separating-in concept and subsequently in practice, once the deal went into effect-the rights of Turkish citizens and those of other citizens in Turkey (that is, potential refugees) allowed a way out of this impasse. In making Turkey 'safe' for refugees to be returned from Greece (even if not for its own citizens), a solution was found. 55 It is important to note, however, that the 'deal' does not actually officially exist as a formal and legally binding agreement between the EU and Turkey. To date, it is based simply on a press release, dated 18 March 2016, stating that 'Turkey and the European Union reconfirmed their commitment to the implementation of their joint action plan activated on 29 November 2015'. 56 Indeed, efforts by NGOs and EU transparency watchdogs like Access Info have confirmed that, beyond the Press Release 144/16 of 18 March, no formal written agreement exists. This ambiguity was put into stark relief by the case brought to the European Court of Justice (ECJ) in February 2017 by an Afghan refugee threatened with expulsion from Greece: the ECJ rejected the case outright, noting that the 'deal' was not an agreement entered into by the European Council or any other EU institution. 57 Despite this, however, the not-really-a-deal deal has made possible very real practices of removal and push-back, and continues to be taken 'as real' by EU institutional actors. In December 2017, EU Commissioner for Humanitarian Aid Crisis Management Christos Stylianides described it as 'one of the biggest humanitarian projects in the world', noting how 'Turkey and the EU together can be proud of this achievement'. 58 From the Turkish side, Foreign Minister Mevlüt Çavuşoğlu has also continued to use the phrase 'refugee deal' (göç anlaşması) while referring to the press statement of 18 March 2016, clearly differentiating it from the readmission agreement (geri the removal of an applicant to a country where he/she risks to be subject to torture and cruel, inhuman or degrading treatment or punishment, as defined in international law, e. the possibility to apply for refugee status exists and, if the applicant is recognized as a refugee, to receive protection in accordance with the Geneva Convention, and f. the applicant has a connection with that country, under which it would be reasonable for the applicant to move to it.' 54 'Turkey as a "safe third country for Greece"', 1. 55 Ibid., 2. In the document, the European Stability Initiative explains that Greece is obliged by EU law and the UN Refugee Convention to admit refugees seeking asylum and investigate their claim. The document suggests, however, that Greek and EU legislation enables Greece to declare an asylum claim inadmissible, return the asylum-seeker to a third safe country and start a 'substantial examination of the claim'. kabul anlaşması), indicating that it would be entirely up to the Turkish government whether or not the deal would remain implemented. 59 As noted previously, the background document preparing the deal suggested that Turkey, by and large, met all necessary criteria of protection, though needed to fully implement the Law on Foreigners and International Protection that grants international protection to asylum-seekers. In practice, however, Turkey entirely lacks the infrastructure and resources to implement this crucial law: yet another slippage consented by the formal ambiguity of the deal. Amnesty International has revealed that refugees in Turkey, of Syrian and other nationalities, regularly face illegal detention, ill treatment in detention and even deportation: either forcibly, or 'indirectly' by being threatened with either return to their country of origin or indefinite detention. 60 Amnesty was not the only organization to raise alarm at the suggested architecture of the deal. The forcible geographical differentiation between the legal definitions and provision of refugee protection garnered the direct condemnation of all major humanitarian agencies. UN High Commissioner for Refugees Filippo Grandi was scathing in his criticism of the provisions of the 'deal', noting that such spurious distinctions (with no basis in legal treaties) risked sending refugees to another country without 'spelling out the refugee protection safeguards under international law'. 61 The UNHCR also expressed its concern at the possibility of the deal's practical implementation before sufficient capacity was provided to Greece to process asylum requests and accommodate refugees pending the examination of their claims. 62 The situation of Syrian minors in Turkey is just one telling example of why Turkey can in no sense be considered a 'safe third country', as the conditions for minors do not provide even the most basic requisites binding under EU law, as indicated in the EU Charter of Fundamental Rights. 63 UNHCR reports that, of the approximately 3.5 million refugees Turkey had taken in from Syria at the time of writing, just over 45 per cent were under the age of 18 and more than 37 per cent were no older than 12. 64 Human Rights Watch, in a report from March 2016, states that their estimates indicate that more than 400,000 Syrian refugee children have no access to education. Indications of UNICEF are even higher: around 80 per cent of Syrian refugees residing in Turkey live outside refugee camps, while only 27 per cent of the children aged between 6 and 17 who live outside these camps have access to education. 65 The Syria Relief Network, an umbrella organization of Syrian humanitarian NGOs, has moreover indicated that many of these children are involved in child labour. 66 The revelations of widespread sexual abuse of minors in Turkish reception centres have provoked further outcry from humanitarian organizations. 67 Amnesty International and other organizations have also documented repeated instances of Turkey engaging in its own push-back operations, including reports of Syrian refugees forced back across the Turkish-Syrian borders en masse since January 2016 by Turkish authorities. Syrian refugees have also been repeatedly denied registration, thus denying them access to basic services, such as medical treatment. 68 The regular push-back of refugees attempting to enter Turkish territory has moreover increased illegal smuggling operations at the Turkish border with Syria, described by Amnesty as 'Fortress Turkey', with hundreds of thousands of displaced Syrians encamped within 20 kilometres of the Turkish border without access to even the most basic facilities.
The Turkish authorities, meanwhile, have begun to put forward proposals to replicate EU attempts to off-shore refugee reception outside its own territory as well. Already in late 2015 Turkey expressed its desire to set up a 'safe zone' for Syrians inside the Syrian borders, in order to avoid taking in more Syrian refugees. 69 Such attempts gained impetus in early 2016 with the progress of the EU-Turkey talks. The suggestion of President Erdoğan that 'cleansing the region of all threatening elements and establishing a safe zone constitutes the basis of 1.7 million Syrian refugees' return' is therefore of grave concern, combined with the current legal status of Syrian refugees inside Turkey that remains ambivalent to say the least: refugees are presently registered as 'guests' and 'brothers and sisters', empty terms in international law. The lack of protection for Syrians and other refugee populations within Turkish territory is further exacerbated by the situation of increasing insecurity and violence in large parts of the country's Southeast, with what many commentators are beginning to refer to as the 'Syrianization' of this region of important PKK presence. 70 Saving Europe?
The evident disconnect in the concerns of EU political leaders between the need to assure certain rights of EU citizens at home-while removing the EU's responsibility in promoting such rights abroad-has long been an object of critique. A large scholarly literature exists on the disjuncture between the EU's rhetoric of rights promotion in its partner countries and actual practice and policies, most importantly in its immediate neighbourhood under the auspices of the European Neighbourhood Policy (ENP) and associated programmes. 71 One of the most potent critiques levelled both by scholars and humanitarian organizations against the ENP in recent years has been its progressive transformation from an initiative intended to promote democratic transformation within partner states, to concerns with securitization. Elspeth Guild wrote about this shift almost a decade ago, noting how the EU's neighbours, especially in the Mediterranean, have now become primarily Europe's police. The Renewed ENP, launched in the autumn of 2015, has further strengthened the primary emphasis on the securitization and bounding of the neighbourhood. Romano Prodi's infamous 'ring of friends' around the EU is now, as The Economist termed it in the summer of 2015, 'a ring of fire'. 72 The EU's neighbourhood is no longer 'a space of opportunity' (whether for economic integration or, in more optimistic times, for the advancement of rights and values that the EU claims to hold dear), 73 but predominantly a space of dangers to be contained, and whose possible 'spill-over' to EU shores and territory must be prevented at all costs. 74 There are two, equally troubling, geographical determinisms at work in such envisionings (and their attendant policy prescriptions and interventions), and two forms of the sort of divisive spatial 'sorting' that we mention at the outset of this article. The first places (in the most literal, geographic sense of the word) the causes of the periodic crises of migration (of which the 2015 refugee crisis is just the most recent incarnation) beyond EU territory and responsibility. As Nicholas de Genova has argued, such understandings map a 'Europe' that is confronted by crises originating beyond its borders; 75 'a "Europe" that is victim of unfathomable conflicts erupting elsewhere, derived from the incapacity or incompetence of (postcolonial) "others" to adequately govern themselves', 76 directly blaming the victims for the wars and violence that are rendering the neighbourhood so 'unsafe'. With this offshoring of blame, the responsibility for securing the region necessarily shifts also-if not predominantly-to the EU's neighbours.
The second form of geographical 'sorting' is instead being directly waged on the bodies of those fleeing the 'ring of fire'. This form of geographical determinism proceeds through a fully colonial desubjectification of migrants, whose decision-making autonomy is entirely denied, as a number of recent studies have argued. 77 The predominant framing evident in most recent EU